So I have been doing HVAC for years as an occupation and only started learning electrical engineering recently. I never knew about transistors so now I am questioning conventional methods a lot. Can't I use MOSFETS? Like contactors for example in an HVAC system. I would love a solid state thermostat for high amp circuits. For instance my 120V window units main board just broke so I wired it direct so it stays on when plugged in. It has an overload on the compressor so it still cycles on/off for safety. I wanted temperature control though so just with any central air system I screwed a contactor and transformer to the wall along with a thermostat. I have used thermostats in the past that just plug into a 120V outlet but never looked inside them to see how they worked because I wasn't as curious back then. It was kneat how they cramed everything into such a small area. I have seen these recently also.
https://www.ebay.com/itm/50-110-C-B...re-Control-Switch-Sensor-DC-12V-/401298480193
All this is so much fun because I have never done anything but walk into a store and buy prebuilt parts and doing everything the way I was taught. I have learned about arduinos and have some coming in the mail. I ordered the Arduino UNO and some shields to play with. I'm going to find some schematics to play with. I want a thermostat that has a 5 degree difference in cut on/off so I figured programming that on an Arduino would be easier for now. I do have some basic programming experience from making video games as a hobby. I would appreciate a point in the right direction on this. Even if it doesn't involve the Arduino. I certainly want to learn the basic of barebones hardware also. Honestly more so. Programming seems very artificial. I'm mainly just wanting to learn how this could be done in a solid state way. I know the link of the item I listed uses a relay but even then I just wonder why do we use contactors and transformers? Thanks for any advice! I had a lot of fun thinking about this.

 

Solid-State-Relays are quite common, but kinda expensive.
The bonus is, they don't wear-out, don't make any noises,
and can be controlled with very little Power.
You basically just need enough Power to light-up a single LED, which isn't much,
( around ~20ma @ ~3-Volts DC ).

You can also make your own SSR using 2 very stout MOSFET Transistors,
if you want to try making stuff yourself.

Since You're normally only switching 1 of the 2 Hot-Legs,
only 1-SSR, or 2 "Back-to-Back" FETs are required.

Here's how to do it with FETs ..........
For 230-Volt operation, some higher-Voltage-rated FETs will be required.
The Schematic also doesn't show a Current-Limiting-Resistor, or, optionally, a proper Current-Regulator for the LED.
The LED must be fed DC-only, which may require the addition of a Diode, and a small Capacitor,
this will depend upon how, and with what kind of Power, You intend to use for your T'stat-Circuit.
A mechanical T'stat is perfectly adequate, trying to control the temperature to very tight tolerances
gets to be complicated very quickly, for this reason, I don't recommend using a Micro-Controller
unless You want to have the Temps change for different times of the Day.
( P.S., this doesn't save much overall power at all, because the "catch-up-period",
when the Temps are commanded back-down,
uses-up all the Power You just saved by raising the Temps earlier ).
.
.
.

 

As @LowQCab suggests, MOSFETs (and Thyristors & IGBTs in related applications) are often used to switch large loads, like everything they have advantages and disadvantages. The reason relays are still around is that they just plain work, are low cost, and can be positively inspected for proper operation since they are in effect just wires with some fancy skills.

@LowQCab’s circuit is a great place to start and I will add that getting a few different SSR (Solid State Relay) modules will help you understand how they are supposed to work while you are trying to cobble together your own. In addition logic level MOSFETs are an option in some applications. These are MOSFETs that will fully switch on with ~5V on the gate. Note that the requirements for AC operation (as in your example) and DC operation are different, but the principles are the same.

The SSR modules that offer logic level operation are going to be such a device or polar of devices with an optocoupler between the input and the relay for galvanic isolation, that is, no electrical connection between the input and the circuit that supplies the power, just light. You can make this yourself as well.

 

There has been 3ph soft-start electronic starters for many years now, I always assumed they used Triacs for the control devices.

 

There has been 3ph soft-start electronic starters for many years now, I always assumed they used Triacs for the control devices.

Forget to mention triacs!

 

For an air conditioner compressor I prefer the magnetic mechanical relay breaking both of thr 230 volt leads to the compressor. The reason is that it provides some actual isolation between the mains feed and the compressor windings. The mains feed sometimes has high voltage spikes that can cause an insulation breakdown, resulting in an expensive compressor replacement being required.
I saw just exactly that this past summer. FOUR compressors suffered a breakdown due to a high voltage spike. That were not running at the time, but to cut their cost, the manufacturer of the very expensive rooftop AC packages used a two pole starter instead of a three pole starter, as it was 3-phase powered compressors. These were 7.5 HP compressors, two per each rooftop unit. The compressors failed because the voltage spike arced between the winding and the laminations, because the one phase was still connected.
This was aterrible act of poor engineering, cutting quality deeply to save a veryy small amount of money. An adequately rated mosfet woulf probably break down with a higher voltage spike like that, an open contact might not flash over. So I would use the mosfets for speed control but still have the starter contacts for OFF/ON operation.

 

I agree that leaving out a set of Contacts to save ~$0.12 cents is kinda cheezy,
and makes You wonder what else they probably scrimped on,
but I have a different take on the Spikes.
Every Distribution-Panel, and every Machine/Unit, should have a
Spike-Suppressor installed when it is first wired-up.
Having multiple Spike-Suppressors installed at every reasonable opportunity will also
tend to add together, and reinforce each other,
and become even more effective at protecting everything else on the property.

Big Roof-top AC-Units are especially vulnerable to Lightning-Strikes.
Nothing will provide protection from a direct Lightning-Strike.
.
.
.

 

I agree that leaving out a set of Contacts to save ~$0.12 cents is kinda cheezy,
and makes You wonder what else they probably scrimped on,
but I have a different take on the Spikes.
Every Distribution-Panel, and every Machine/Unit, should have a
Spike-Suppressor installed when it is first wired-up.
Having multiple Spike-Suppressors installed at every reasonable opportunity will also
tend to add together, and reinforce each other,
and become even more effective at protecting everything else on the property.

Big Roof-top AC-Units are especially vulnerable to Lightning-Strikes.
Nothing will provide protection from a direct Lightning-Strike.
.
.The units damaged were rooftop units, but between peaked roofs much higher. So it might have been lightning, but my bet is on a spike, because there was damage to a computer inside the building. The power supply was destroyed.
.